aan

de

rtmeical

form

22

sectors and produces 50e100 L wastewater/kg of finished

anoxic)-based processes are widely used for textile wastewater

though some of the more biodegradable dye auxiliaries may

treatment systems cannot achieve destructive decolorizationdue to the fact that textile dyes are intentionally designed to

(being a mixture of H2O2 and Fe2 and applied at acidic pH)

can be effectively used to achieve complete color and partialCOD removal from textile effluent and thus an attractive optionto prepare recalcitrant process streams to conventional acti-vated sludge treatment of the combined wastewater. However,

* Corresponding author. Tel.: 90 212 285 37 86; fax: 90 212 285 65 45.E-mail address: arslanid@itu.edu.tr (I. Arslan-Alaton).

Available online at www.sciencedirect.com

Dyes and Pigments 78 (200product [1]. From the environmental point of view, particu-larly the textile dyeing process constitutes a major pollutionproblem due to the variety and complexity of chemicals(dyes, sequestering agents, tannins, dye carriers, levelingagents, dispersing agents, etc.) employed. Among the textileauxiliaries, dyes have attracted the most attention since colorin dyehouse effluent not only causes environmental concerns,but also creates a significant aesthetic problem in sewage treat-ment works and receiving water bodies [2e4].

Conventional chemical (coagulationeflocculation) and bio-logical (activated sludge, sequential bed reactors, anaerobic/

resist biological, photolytic and chemical degradation. The na-ture of textile effluent depends on fashion, technical, techno-logical, social and economical factors. These effluents oftenrequire pre-treatment of segregated process streams (e.g. dye-bath effluent) using alternative, advanced oxidation processes(AOPs) that have more recently been used to treat refractoryand/or toxic pollutants [7e9].

Among the AOPs, the Fentons process is quite well knownand has been successfully applied for the treatment of dyehouseeffluent [10e17]. The Fentons reagent is relatively cheap andeasy to handle compared with other AOPs. Fentons reagentThe textile preparation, dyeing and finishing industry is oneof the mightiest water consumers among different industrialbe completely eliminated from dyehouse effluent, conventionalThe effect of untreated and Fenton-treated acid dyes (C.I. Acid Red 183 and C.I. Acid Orange 51) and a reactive dye (C.I. Reactive Blue 4) onaerobic, anoxic and anaerobic processes was investigated. The optimum Fe2:H2O2 molar ratio was selected as 1:5 (4 mM:20 mM) for 10 minFenton treatment at pH 3, resulting in reduced chemical oxygen demand and dissolved organic carbon removal efficiencies; only acetate wasdetected as a stable dye oxidation end product. During anaerobic digestion, 100, 29% and no inhibition in methane production was observed forthe untreated blue, red and orange dyes, respectively. The inhibitory effect of the blue reactive dye on methane production was w21% afterFenton treatment. Neither untreated nor treated dyes exhibited an inhibitory effect on denitrification. Aerobic glucose degradation was inhibitedby 23e29% by untreated dyes, whereas Fenton-treated dyes had no inhibitory effect on aerobic glucose degradation. 2007 Elsevier Ltd. All rights reserved.

Keywords: Dyehouse effluent; Acid dyes; Reactive dyes; Fenton treatment; Aerobic, anoxic and aerobic treatment processes

1. Introduction treatment, however, with a rather limited success [5,6]. Al-Advanced oxidation of acidFenton treatment on aerobic,

Idil Arslan-Alaton a,*, Betul Hana Istanbul Technical University, Faculty of Civil Engineering, Depa

b Institute of Environment and Resources, Techn

Received 18 September 2007; received in revised

Available online

Abstract0143-7208/$ - see front matter 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.dyepig.2007.11.001nd reactive dyes: Effect ofoxic and anaerobic processes

Gursoy a, Jens-Ejbye Schmidt b

nt of Environmental Engineering, 34469 Maslak, Istanbul, TurkeyUniversity of Denmark, 2800 Lyngby, Denmark

1 November 2007; accepted 5 November 2007

November 2007

8) 117e130www.elsevier.com/locate/dyepig

advanced oxidation products might be more toxic and/or inhib-itory on the biological treatment systems used for the post-treatment than the original textile dyes. In fact, many studieshave recently reported the effect of Fenton pre-treatment oftextile (mainly reactive and acid) dyes on aerobic biologicaltreatment processes by coupling the Fenton (or Photo-Fenton)process with an aerobic biological (mainly sequential batch) re-actor; however, no study dealing with the influence of Fentonoxidation of textile dyes on anaerobic and anoxic biologicalprocess has been published so far [17e22]. Due to the factthat the textile dyer and finisher is mainly confronted with an-aerobic and anoxic treatment systems, it is more interesting andimportant to study the behavior of these dyes and their degra-dation products under anaerobic and anoxic conditions.

Considering the above mentioned facts, the aim of thepresent study was to evaluate the effect of Fentons treatmentof three commercially important textile industry dyes onconventional biological (anaerobic, anoxic and aerobic) pro-cesses. For this purpose, two acid dyes, namely Acid Red183 (AR 183) and Acid Orange 51 (AO 51), and one reactive

2. Materials and methods

2.1. Textile dyes

Two acid dyes (AR 183 and AO 51) and one reactive tex-tile dye (RB 4) were chosen as refractory model pollutantsand all purchased from Aldrich. Particularly these three textiledyes were selected for this study since their molecular struc-ture is exactly known and all three dyes are frequently beingapplied for the dyeing of cotton, woolen and nylon (polyam-ide) fabrics worldwide as well as in Turkey. Some importantphysicochemical properties of the selected textile dyes arepresented in Table 1. For the Fenton treatment as well as an-aerobic, anoxic and aerobic experiments, 100 mg/L aqueousdye solutions (corresponding to 171 mM AR 183, 116 mMAO 51 and 157 mM RB 4) were prepared in distilled water.The reason why particularly 100 mg/L was selected in thepresent study are formerly published papers [17,21,23] aswell as private communications with technical staff from localdyehouses [24e26] indicating that the dye concentration typ-

O 5

36H

60.8

6,55

63

118 I. Arslan-Alaton et al. / Dyes and Pigments 78 (2008) 117e130dye, i.e. Reactive Blue 4 (RB 4), were selected as model tex-tile dyes and subjected to Fentons treatment under differentoxidant (H2O2) and transition metal catalyst (Fe

2) concen-trations. Fenton treatment efficiency was assessed in termsof color, COD (chemical oxygen demand) and DOC (dis-solved organic cabon) removal rates. The study also aimedat quantitatively identifying stable advanced oxidation endproducts that were expected from textile dye degradationsuch as carboxylic acids (formic, acetic, malic, maleic, etc.)as well as inorganic salts such as sulfate, nitrite, nitrate andchloride. In the second part of the study, the inhibitory effectof untreated and Fenton-treated textile dyes on biologicalprocesses was examined in terms of (a) methane production,(b) denitrification and (c) glucose-COD activated sludgetreatment rates.

Table 1

Physicochemical properties of the textile dyes AR 183, AO 51 and RB 4

Textile dye AR 183 A

Molecular formula C16H11ClN4Na2O8S2$xCr CMolecular weight (g/mol) 584.84 8

C.I. number 18,800 2

lmax (nm) 497 4

Molecular structureically encountered in effluents originating from the cotton andpolyamide dyeing factories is in the range of 10e200 mg/L.Environmental characterization of 100 mg/L aqueous solu-tions of the selected dyes in terms of COD, DOC and absor-bance parameters is given in Table 2. The other reagents usedin the present study were hydrogen peroxide (H2O2; 35% w/w,Fluka), ferrous iron sulfate heptahydrate (Fe(SO4)$7H2O,Fluka), and enzyme Catalase (from Micrococcus lyseidicticus;1 AU destroys 1 mmol H2O2 at pH 7 RTP, 100,181 U/mL,Fluka) to destroy residual, unreacted H2O2. Peroxide Quanttest strips (Aldrich) were used to determine the approximateamount of residual (unreacted) H2O2 in the Fenton reactionsolutions. Millipore syringe filters with 0.45 mm cutoffswere used to separate supernatant after ferric hydroxide pre-cipitation and to cease the Fenton reaction.

1 RB 4

26N6Na2O11S3 C23H14Cl2N6O8S20 637.43

0 2,363,639

595

All Fenton experiments were run with 1000 mL, 100 mg/L3 in

1500 mL capacity glass beakers. Magnetic stirrers and stir

Anaerobic experiments were conducted with untreated dyesitial

with the exact amount of daily prepared, diluted Catalase solu-o ex-

amine the effect of the Catalase enzyme on anaerobic biomass.

ndas well as dyes being subjected to Fentons reagent at an in2bars were used to provide continuous mixing during theFenton experiments. In order to initiate the Fenton reaction,appropriate amounts of H2O2 stock solution (10.29 M) toachieve the desired H2O2 concentration in the final reactionsolution, and likewise, Fe(SO4)$7H2O from a 10% w/v(0.36 M) stock solution to achieve the desired Fe2 concentra-tion, were added to the dye solution simultaneously and rightafter pH adjustment with 0.02, 0.1, 0.5 and 1.0 N H2SO4 solu-tions. Twenty five milliliters of sample aliquots were takenfrom the beakers at time intervals of 0, 2, 5, 10, 20 and30 min during the Fenton reaction for analyses. For each sam-ple, the reaction was quenched immediately with the additionof 6 N NaOH solution to increase the pH to around 9e10. An-other, second adjustment was done with 0.01 and 0.5 N H2SO4to pH 7e8 in order to achieve the maximum amount ofFe(OH)3 precipitation. After Fe

2 was practically completelyprecipitated out of the reaction solution in the form ofFe(OH)3 (ferric hydroxide) sludge, the sample was filteredthrough 0.45 mm-cutoff Millipore filters and spiked with suffi-cient amounts of enzyme Catalase to destroy any residual/unreacted H2O2 in order to prevent its positive interferencewith COD measurements. Catalase also contributed to theCOD (1.46 mg COD/mL Catalase) and DOC (1.02 mg DOC/mL Catalase) of the treated samples as determined by prepara-tion of Catalase control samples (e.g. H2O2 added todistilled water and mixed with exactly the same amount ofCatalase that was used to destroy H2O2 in the reaction sam-ples). Fenton-treated samples were subjected to color, CODand DOC analyses after filtration.

2.3. Anaerobic experimentsaqueous dye solutions at T 20 C and an initial pH of2.2. Fenton treatment

Table 2

Environmental characterization of 100 mg/L textile dye solutions AR 183

(171 mM), AO 51 (116 mM) and RB 4 (157 mM)

Textile dye AR 183 AO 51 RB 4

COD (mg O2/L) 50 103 63

DOC (mg C/L) 14 31 21

Absorbance at lmax (cm1) 1.02 at 497 nm 1.75 at 463 nm 0.68 at 595 nm

I. Arslan-Alaton et al. / Dyes apH of 3; an Fe :H2O2 ratio of 1:5 (4 mM:20 mM) and a treat-ment time of 10 min, for 60 days at T 37 C (under meso-philic conditions). The bioreactors were run in triplicate foreach untreated and Fenton-treated dye sample. The pH of thedye samples and the anaerobic biomass were adjusted to 7.5and 6.8, respectively. Before the anaerobic experiments wereinitiated, oxygen was completely removed from the systemsby purging the bioreactors with a mixture of 80% N2 20%CO2 (v/v). The biomass (determined as 16,200 mg/L in termsof VSS (volatile suspended solids)) used in the anaerobicThe VSS concentration in all bioreactors was adjusted to500 mg/L. No additional external carbon source was used inthe bioreactors to observe the effect of the untreated and treateddyes on the methanogenic activity more clearly. However, nomethane production was observed within the first 48 h of theanaerobic experiments in all batch reactors including controls.For this reason, 500 mg/L glucose solution was added to one ofthe control bioreactors to observe if the delay in methaneformation was related to the absence of readily degradablesubstrate. Samples were taken at regular time intervals up to60 days for color, COD, DOC, VSS and TSS analyses.

2.4. Anoxic experiments

Anoxic experiments were conducted with untreated dyes aswell as dyes being subjected to Fentons treatment (pHo 3;Fe2:H2O2 4 mM:20 mM; t 10 min) for 48 h at T20 C. For the anoxic experiments, sludge (VSS 6400 mg/L) was collected from the anoxic/aerated tank after theprimary settling tank of Lundtofte Urban Wastewater Plant(Biodenitro plant) located in Lyngby, Denmark. The pHof the untreated/treated dye samples and the biosludge wereinitially adjusted to 7.5 and 6.8, respectively. Sludge was notadapted to the dyes and Fenton oxidation products. At the be-ginning of the denitrification experiments, NaNO3 (Fluka) andNaCH3COO (Riedel) were added to the reactors at concentra-tions of 111.8 mg/L (1.31 mM) and 335.6 mg/L (5.69 mM),respectively. The VSS concentration in all bioreactors was se-lected as 2400 mg/L. The anoxic batch reactors (one typicalreactor is depicted schematically in Fig. 1(b)) consisted of14, 2000 mL capacity glass bottles (total sample volu-me 1600 mL) from which sample aliquots were taken everyhalf an hour during the first 4 h of the experiment and thention to destroy exactly 20 mM H2O2, were also prepared texperiments was obtained from the anaerobic digestor of Lund-tofte Urban Wastewater Plant located in Lyngby, Denmark,treating a mixture of primary and secondary sludge. The anaer-obic sludge neither was adapted to the dyes nor the Fenton ox-idation products both of which consequently served as the solecarbon source. The anaerobic batch reactors schematically pre-sented in Fig. 1(a) consisted of 56, 500e2000 mL capacityglass bottles out of which 10 (total sample volume 1600 mL)were used for regular color, COD, DOC, VSS and TSS (totalsuspended solids) analyses and the remaining ones (total sam-ple volume 300 mL) were kept for regular CH4 analyses.Formalin (37% w/v, Merck) was added at a concentration of2% (w/v) to the abiotic control reactors. In addition, control(containing inoculum and distilled water only) and Catalasecontrol (containing inoculum, distilled water, H2O2 and en-zyme Catalase) reactors were prepared. The control reactorsrefer to the samples that neither bore untreated nor Fenton-treated textile dyes but only water and inoculum. Catalasecontrol reactors, containing 20 mM H2O2 solution spiked

119Pigments 78 (2008) 117e130after 24 and 48 h for color and nitrate analyses. Pure argon(100% v/v) was used to remove nitrogen, oxygen and other

r co

)

nd(I) Schematic of anaerobic reactor fo

VSS-TSS analysis

Vsample = 1600 mL

(I(3)

(2)

a

120 I. Arslan-Alaton et al. / Dyes agases in the bioreactors before the experiments were started.Formalin (37% w/v) was added to abiotic reactors at a concen-tration of 2% v/v. In addition, separate control and Catalasecontrol reactors were run in parallel as described for the anaer-obic experiments.

2.5. Aerobic experiments

Aerobic experiments were conducted separately for un-treated dyes and Fenton-treated oxidation intermediates to ex-amine their effect on glucose degradation with heterotrophicbiomass. For that purpose, activated sludge experiments wererun for 6 h in 8, 1000 mL fed-batch reactors at T 20 C andpH 7.0e7.5 with mixed bioculture that was obtained fromthe aeration tanks of Tuzla and Baltaliman Urban WastewaterTreatment Plants located in Istanbul, Turkey. The activatedsludge was first acclimated to glucose solutions (CODo (initialCOD) 500 mg/L) for 8 weeks and thereafter used in the aer-obic experiments. Samples containing 516 mg/L glucose(500 mg/L COD) and 100 mg/L untreated textile dyes or

(II) Schematic of anaerobic reactor for C

Vsample = 1600 mL

(3)

(2)

(1)b

Fig. 1. Schematic representation of the (a) anaerobic and (b) anoxic experimental se

valve to maintain a closed system; (3) tubing providing the connection between salor, COD, DOC,

(II)

Vsample = 300 mL

(1)(1)Pigments 78 (2008) 117e130textile dyes being subjected to Fenton treatment (pHo 3;Fe2:H2O2 4 mM:20 mM; t 10 min) were individuallyprepared together with their corresponding controls (516 mg/L glucose only (500 mg/L COD), appropriate pH buffersand macro/micronutrients. The VSS concentration was keptat 4400 mg/L for the experiments with untreated and3700 mg/L for the experiments with Fenton-treated dyes tomaintain the F/M (food-to-microorganism) ratio in the aerobicsample and control bioreactors at 0.12 and 0.13 mg CODo/mgVSS, respectively. During the aerobic experiments, sample ali-quots were taken at periodic time intervals up to 6 h for color(absorbance), COD, TSS and VSS analyses after filtrationthrough 1.20 (Whatmann) and 0.45 (Millipore) mm-cutoff fil-ters. Again, control and Catalase control reactors were preparedas described for the anaerobic as well as anoxic experiments.

2.6. Analytical tools

In this study, color was referred to as the absorbance mea-sured at the maximum absorption bands of the investigated

H4

analysis

t-ups. Elements of the anaerobic and anoxic reactors: (1) Sampling syringe; (2)

mple and syringe.

expected. However, the improvement in treatment efficienciesidantf the

other reagent (2 mM in the case of Fe and 20 mM for

ndtextile dyes (i.e. 497 nm for AR 183, 463 nm for AO 51 and595 nm for RB 4). The absorbance of the untreated and treatedsamples was measured on a Jenway 6400 model UVevis(ultravioletevisible) spectrophotometer at the Technical Uni-versity of Denmark (DTU), Denmark, and on a Novespec II/Pharmacia LKB colorimeter at Istanbul Technical University(ITU), Turkey. Absorbance measurements were performedwith 1 cm optical path length, reusable glass cells. COD mea-surements were based on Danish Standard (DS) No. 217 [27]for low-range (15e100 mg O2/L) COD at DTU, Denmark,whereas COD measurements were conducted in accordancewith ISO 6060 [28] by the closed reflux, titrimetric methodat ITU, Turkey. DOC was measured on a Shimadzu TC (totalCarbon) 5000A carbon analyzer after filtration of the samplesthrough 0.45 mm Millipore syringe filters. In order to calibratethe instrument, several dilutions of potassium hydrogen phtha-late and potassium bicarbonate solutions were prepared usingcertified references for TC and IC (Inorganic Carbon) analyses(500 mg/L each, QC WW4A Eurofins). An ICS-1500 modelDionex ion chromatograph equipped with an Ion Pac AS14 mm (10e31) Column (P/N 46124) in combination withan anion suppressor (AMMS III 4-mm P/N 56750) was usedto measure nitrite, nitrate and sulfate and carboxylic acids inthe untreated and treated samples during Fenton and anoxicexperiments. Column and cell heater temperatures were setas 35 C, whereas the column pressure was selected as1400 psi. A Shimadzu model-14A gas chromatograph equip-ped with a Porapak 60/80 Molsiere column (6 ft in lengthand 3 mm in inner diameter) and a flame ionization detector(FID) was used for CH4 measurements during the anaerobicexperiments. Nitrogen (100% v/v) was used as the carriergas at a pressure of 2.0 kg/cm2. The injection temperaturewas set as T 110 C, while the detector and oven tempera-tures were adjusted to 160 C. The retention time for CH4was around 50 s. A GC Shimadzu GC-2010 equipped witha capillary column (ZB-FFAP, 30 m 0.53 mm 1.0 mm)and an FID were used to measure acetate during anaerobicand Fenton experiments. TSS and VSS were measured accord-ing to a procedure outlined in Standard Methods [29]. pH wasmeasured using digital pH meters (Metrohm, model 692, atDTU and Thermo Orion, model 520 at ITU). The chromiumcontent of AR 183 before and during Fenton treatment was an-alyzed on a Unicam 929 model atomic absorption spectrome-ter with and without 50% v/v nitric acid digestion to determinethe total amount of chromium and chromium released from theorganically bound trivalent chromium complex azo dye duringFenton oxidation, respectively. Unfortunately, chromiumleached out of the chromium complex acid dye directly afteracidification with 14.5 N HNO3, even before the dye was sub-jected to Fentons treatment. The total chromium content wasdetermined as 4.5 mg/L in the 100 mg/L digested, original dyesample and 3.8 mg/L in the undigested dye sample. Chromiumlevels decreased to 1.2 mg/L for Fenton-treated dyes thatmight be attributable to adsorption of the released chromiumions and/or organically bound chromium on Fe(OH)3 sludge

2/3

I. Arslan-Alaton et al. / Dyes athat was formed during precipitation of Fe ions to ceasethe Fenton reaction and to remove residual soluble iron.H2O2). This was particularly evident for the Cr(III) complexwas more dramatic when either the catalyst or the oxconcentration was doubled at the lower concentration o

23. Results and discussion

3.1. Fenton experiments: establishment of optimumworking conditions

The main process variables affecting the rate of Fentonsreaction are the molar concentrations of the oxidant (H2O2)and catalyst (Fe2), particularly the Fe2:H2O2 molar ratio. In-creasing the H2O2 concentration is important to obtain highoxidation efficiencies, while elevating the Fe2 concentrationdirectly enhances the oxidation rate [11,13,16,30]. However,it should be kept in mind that if the concentration of one reac-tant is increased to observe its positive effect, thereby keepingthe other one constant, the Fe2:H2O2 molar ratio and hencethe oxidation condition will change significantly. When oneof the reagents is provided in excess, a dramatic reduction inthe oxidation efficiency is expected. This can be explainedby the following scavenging reactions of

OH (hydroxyl radi-

cals) [31]:

Fe2 OH/ Fe3 OH k20C 4:3 108 M1 s11

H2O2 OH/H2O HO2 k20C 2:7 107 M1 s12

Generally speaking, advanced oxidation of organic com-pounds is fast when ferrous ion is present at a concentrationvarying between 2 and 5 mM, e.g. a concentration rangewhere sufficient

OH are produced and Fe2 is still highly

soluble in water at pH 2e5. In the present study, a set of pre-liminary (baseline) Fenton experiments was carried out with100 mg/L aqueous AR 183, AO 51 and RB 4 solutions atan initial pH of 3 (i.e. the optimum pH for Fenton reactions)at different Fe2:H2O2 molar ratios selected as 2:20, 2:40,4:20 and 4:40 (in mM). These selected Fe2 and H2O2 con-centrations are corresponding to optimum Fenton molar ratios(1:5 and 1:10). The most suitable molar ratio for each dyewas individually determined upon inspection of the obtainedcolor and CODeDOC removal rates. In order to examinethe extent of degradation, possible nitrite, nitrate, sulfateand carboxylic acid formation was also followed. Table 3summarizes percent COD and DOC removal efficiencies ob-tained after 10 and 30 min Fenton treatment for all the threestudied textile dyes. Since color removal was very fast andpractically complete (>95%) within the first 2 min of the re-action, color abatement was not further considered as a criticalprocess parameter in these baseline experiments. From Table3 it is evident that treatment efficiencies generally speakingincreased with increasing Fe2 and H2O2 concentrations, as

121Pigments 78 (2008) 117e130azo dye AR 183 and the anthraquinone dye RB 4. AO 51was not so seriously affected by the variations in molar

0.0

0.2

0.4

0.6

0 5 10 15 20 25 30

0 5 10 15 20 25 30

Time (min)

No

rm

alized

C

olo

0

20

40

60

80

100

120

Time (min)

CO

D (m

g/L

)

35

AR 183 AO 51 RB 4

b

c

2:20 32 38 28 32

ndconcentrations and ratios of Fe2 and H2O2 implying thateven the lowest tried doses were sufficient for partial oxida-tion of this disazo dye. However, COD and DOC removal ef-ficiencies even started to decrease when the Fe2:H2O2 molarratio was increased from 1:5 to 1:10 for 4 mM Fe2, indicat-ing that H2O2 was overdosed when its concentration wasincreased to 40 mM. From these findings it can also be in-ferred that the most suitable Fe2:H2O2 molar ratio was foundto be 1:5 at Fe2 and H2O2 concentrations of 4 and 20 mM,respectively, coinciding with the highest COD and DOC re-

2:40 50 56 42 49

4:20 56 58 50 54

4:40 55 57 47 50

AO 51

2:20 66 72 57 67

2:40 67 78 64 66

4:20 75 80 75 76

4:40 70 77 70 74

RB 4

2:20 38 38 28 32

2:40 48 56 30 49

4:20 66 58 47 53

4:40 57 57 40 43

a Percent COD and DOC removal after 10 min (t10) and 30 min (t30) Fentontreatment at pHo 3.Table 3

Results obtained for the baseline Fenton experiments conducted with 100 mg/L

aqueous AR 183 (171 mM), AO 51 (116 mM) and RB 4 (157 mM) at different

Fe2:H2O2 molar ratios and an initial pH of 3

Fe2:H2O2 (mM:mM) COD removala (%) DOC removala (%)

t10 t30 t10 t30

AR 183

122 I. Arslan-Alaton et al. / Dyes amoval efficiencies for both 10 and 30 min Fenton treatment.In a related investigation conducted with the Fentons reagent,an optimum molar ratio of 1:4 was established for the treat-ment of simulated acid dyebath effluent bearing azo and an-thraquinone dyes [16]. In another study, the optimum molarratios were determined as 1:5 and 1:7 for the Fenton oxida-tion of Disperse Blue 106 and Disperse Yellow 54, respec-tively [13]. The differences in treatment efficiencies may beattributed to the differences in their molecular structuresand molar concentrations. AO 51 has two azo bonds com-pared to the single azo bond of the chromium complex azodye AR 183 and an appreciably higher molecular weightthan AR 183. The molar concentration of AO 51 is conclu-sively less than that of AR 183.

3.2. Fenton treatment

3.2.1. ColorConsidering the results obtained in the preliminary Fenton

experiments, it was decided to continue the study under thefollowing reaction conditions; initial pH (pHo) 3;Fe2 4 mM and H2O2 20 mM. Fig. 2(a) presents colorabatement rates for the three textile dyes under the above men-tioned working conditions. From Fig. 2(a) it is obvious that0.8

1.0

1.2

r (A

/A

o)

AR 183 AO 51 RB 4

a

Pigments 78 (2008) 117e130color removal was practically complete within the first 2 minof Fenton reaction. For the azo dyes (AR 183 and AO 51)99% and for the reactive dye RB 4 93% color removal wasachieved after 30 min, respectively. The rapid decolorizationrates can be explained by the fast reaction between Fe2 andH2O2 to produce a sufficient amount of

OH that promptly

cleaved the dye chromophores. In other words, the in situformed oxidants (free radicals such as

OH and HO2

) had

0 5 10 15 20 25 300

5

10

15

20

25

30

Time (min)

DO

C (m

g/L

)

AR 183 AO 51 RB 4

Fig. 2. (a) Normalized color, (b) COD and (c) DOC abatement during Fenton

treatment of 100 mg/L AR 183, AO 51 and RB 5 (Fe2 4 mM;H2O2 20 mM; pHo 3).

ndenough oxidation capability to degrade the chromophores ofall of the studied dyes at a high degree (>90%). From previousstudies reported in the literature, fast and complete color re-moval (>99.9%) was achieved by employing Fentons reagentfor 50 mg/L Acid Orange 7 [21]. Similarly, color removal wasobtained as 96% for 50 mg/L Reactive Black 5 during the firstminute of Fentons reaction [17]. As is also evident fromFig. 2(a), color removal proceeded so fast that is was not pos-sible to fit the color abatement curve to any reaction kinetic.

Fast color removal was accompanied with a parallel, rapiddecrease in H2O2 concentration (only 40% of the initial H2O2remained in the reaction solution after 5 min Fenton treatmentand was almost completely consumed at the end of the treat-ment process) indicating rapid consumption of the Fenton re-agent as the oxidation proceeded. Indeed, fast consumption ofH2O2 during the early stages of textile dye degradation viaFenton treatment has been reported before by many re-searchers [21,32,33].

3.2.2. CODFig. 2(b) displays changes in COD during Fenton treatment

of the studied textile dyes. From Fig. 2(b) it is evident thatCOD abatement started right after initiation of the reactionand slowed down after only 5 min treatment most probablydue to the accumulation of advanced oxidation intermediatesthat are more resistant to further oxidation than the textiledyes, as well as completion of the decolorization process. Inall cases, only partial COD removal was achieved at the endof the reaction due to the fact that after cleavage of the dyechromophores the reaction slows down such that the highlycomplex-structured dye molecules are only partially degradedto relatively small organic fragments, such as carboxylic acids,aldehydes, ketones and alcohols [34,35]. For this reason, com-plete oxidation (mineralization) of the dyes was not expected.This observation can be supported by a former, related workwhere Fenton oxidation of synthetic acid dyebath effluent re-sulted in only 23% COD removal after 30 min treatment [16].In another investigation which was carried out with real, com-bined textile wastewater, only 30% COD removal could beachieved after Fenton treatment. This behavior was attributedto the relatively short retention time selected for the treatmentprocess and low reagent doses leading to incomplete degrada-tion and accumulation of stable oxidation intermediates in thereaction medium [12].

3.2.3. DOCDOC abatement of the three textile dyes was also followed

under the Fenton treatment conditions and shown in Fig. 2(c).From Fig. 2(c) it is apparent that results obtained for DOCand COD removals were almost parallel to each other; DOCabatement proceeded very fast and showed an asymptotic be-havior during the first 5 min of the reaction, speculatively dueto accumulation of oxidation intermediates and consumptionof the Fenton reagents. As valid for color and COD removalrates, the decreasing order of DOC abatement rates was ob-

I. Arslan-Alaton et al. / Dyes atained as AO>AR> RB. This observation is a consequenceof differences in molecular weights and structures; it is knownthat the azo bond is more prone to oxidative attack byOH

than the anthraquinone-based chromophoric grouping [5].

3.2.4. Identification of final (stable) oxidation productsMechanistic studies were mainly devoted to the ozonation

of azo dyes and their number increased recently mainly be-cause case studies dealing with the effect of advanced oxida-tion processes on the toxicity of textile dyes demonstratedthat toxicity may increase if treatment conditions are not opti-mized and/or treatment time is kept too short. Analytical stud-ies conducted to elucidate the degradation mechanism oftextile dyes have been rather qualitative than quantitative,based on the assumption that

OH play a major role in the

degradation pathway. Former studies have indicated thatOH-addition to N-positions (in the azo bond) and differentC-positions (in the phenol and naphthol rings), that, with thehelp of the hydrogen radical, were subsequently cleaved to in-termediates such as 2-hydroxynaphthaquinone, hydroquinone,benzo-1,4-quinone, naphthalene-2-sulphonic acid, urea, 1-naphthol and acetamide, is the initial step of oxidative dyedegradation. The benzene rings were further cleaved to yieldphthalic acid and/or but-2-enedioic acid and finally degradedto aliphatic carboxylic acids such acetic and formic acid.Most intermediates that are probably involved in the degrada-tion of azo dyes remain tentative states and could not even bequalitatively determined via ion chromatography (IC) and gaschromatography/mass spectrometry (GC/MS) [36]. In anotherstudy it was anticipated that during advanced oxidation ofmetal-complex azo dyes the rupture of azo and metal oxygengroups takes place initially, followed by further oxidation ofthe aromatic compounds to alcohols, aldehydes and carboxylicacids. Released metal ions (including cobalt, copper and chro-mium) could be effectively removed via metal hydroxide pre-cipitation under alkaline conditions. In a recent investigation,GC/MS analysis of the cobalt-complex azo dye Acid Brown159 that was subjected to ozonation and Fenton treatment,resulted in the identification of acetophenone, 1-methyl-1-ben-zimidazol-2-amine, N-(3,4-dimethyl-2,6-dinitrobenzyl) pentan-3-amine and N-methyloaniline [37] as the dye intermediates.On the other hand, initial ozonation products of ReactiveRed 120 were identified as phenol, 1,2-dihydroxysulfobenzeneand 1-hydroxysulfonbenzene via high performance liquidchromatography (HPLC)/mass spectrometry (MS/MS) [38].Final and hence smaller molecular-weight oxidation productswere detected as acetate, sulfate and nitrate in that study asmany reactive dyes, Reactive Red 120 comprises azo, triazineand amino groups. However, in these studies it is believed thatthe nitrogen in the azo bond is converted to nitrogen gas(molecular nitrogen) since nitrate could not be detected oronly in trace amounts during advanced oxidation and ozona-tion of many azo dyes including Reactive Red 120 and CongoRed [39,40]. Harsh oxidation conditions were needed for theoxidation of compounds bearing 1,3,5-triazinyl-groups suchas (amino)chlorotriazine reactive dyes [38,41]. In another re-cent investigation, solid-phase extraction coupled with GC/

123Pigments 78 (2008) 117e130MS analyses was employed to identify dye intermediatesduring advanced oxidation of Reactive Orange 113. Octanal,

acetophenone, nonaldehyde, butylethyl acetic amide, 1,2-ben-zenedicarboxylic acid, diethylester, as well as substituted phe-nols and amides were detected and a possible degradationpathway was postulated. In that experimental study, it wasdemonstrated that naphthalenes were degraded to substitutedbenzenes and alcohols, and substituted benzenes on the otherhand were cleaved to aldehydes and carboxylic acids. Again,the triazine group remained intact during ozonation and theFenton process. Neither ozonation nor Fenton process wascapable of achieving complete mineralization (TOC abatementwas always less than 40%) [42]. The present investigationaimed at quantifying stable oxidation intermediates (low-molecular-weight carboxylic acids such as formic, malic,maleic, acetic acid) and oxidation end products (such assulfate, nitrite, nitrate, chloride, etc.). However, none of thesecompounds with the exception of acetic acid and sulfate wasquantifiable at significant levels. Sulfate results were rathermeaningless since ferrous sulfate was employed as the ferrousiron source for the Fenton experiments and remained constantthroughout the Fenton process, whereas the concentrations ofnitrite and nitrate were always found less than 0.2 mg/L. Nolow-molecular-weight carboxylic acid other than acetic acidcould be detected during Fenton treatment of the three textiledyes. Hence, Fig. 3 only depicts acetate formation (a) and itscontribution to the total DOC (b) remaining from thedegradation of the three textile dyes via Fenton oxidation un-

removal was found 80e100% after 15 h [45], whereas Albu-

124 I. Arslan-Alaton et al. / Dyes andder predetermined reaction conditions. In all cases, acetateconcentrations were less after 30 min as compared with the ac-etate concentration found after 10 min Fenton treatment. This

0

5

10

15

20

0 10 30

0 10 30

Acetate (m

g/L

)

a

0

20

40

60

80

100

Time (min)

Percen

t o

f T

otal D

OC

(%

)

b

AR 183 AO 51 RB 4

Fig. 3. (a) Acetate formation and (b) acetate-DOC given as a fraction of thetotal DOC during Fenton treatment of 100 mg/L AR 183, AO 51 and RB 5

(Fe2 4 mM; H2O2 20 mM; pHo 3).querque et al. [46] observed 90% decolorization of 25 mg/LAcid Orange 7 in the presence of 1150 mg/L starch(1000 mg/L COD) that served as the external carbon source af-ter only 10 h. On the other hand, Maas and Chaudhari [47] whoexamined the anaerobic decolorization of 100 and 200 mg/LReactive Red 2 found that biosorption played a major role incolor abatement and accounted for more than 80% removalof the original dye. The overall decolorization efficiency wasobtained as 78 and 76% after 27 days for 100 and 200 mg/LReactive Red 2, respectively. In another work carried outwith Reactive Blue 5 at a concentration of 100 mg/L, color re-moval was 37% after 16 days [48] and it could be demonstratedthat decolorization was mainly due to biosorption. The absenceof an external co-substrate in the present study could have re-sulted in an inefficient decolorization process via reductiveobservation may be attributable to its gradual conversion toCO2 and H2O as the reaction proceeded. The decreasing orderof acetate formation rate was observed to be inversely propor-tional to the decreasing DOC abatement rates, namely RB4>AR 183>AO 51. This may be explained as follows: Fen-ton oxidation was appreciably faster for AO 51 than for theother two dyes and hence a significant portion of the formedacetate (advanced oxidation intermediate) was already con-verted to CO2 (oxidation end product) and H2O upon exten-sion of the reaction time from 10 to 30 min. Ninety sixpercent of the final DOC originated from the acetate thatwas formed during Fenton oxidation of AR 183. Almost20 mg/L acetate accumulated in the reaction solution whenRB 4 dye was subjected to Fenton oxidation for 10 min anddecreased to 17 mg/L after 30 min treatment time.

3.3. Anaerobic experiments

3.3.1. ColorChanges in color during anaerobic experiments of untreated

AR 183, AO 51 and RB 4 are given in Fig. 4(a). Initial sorptiononto biomass (the absorbance removed before the experimentwas initiated via physical adsorption) accounted for 15, 65and 25% AR 183, AO 51 and RB 4 removals, respectively(not shown data), while by total biosorption (preliminary ad-sorption plus biosorption during the experiment) 46% (AR183), 89% (AO 51) and 66% (RB 4) color removals wereachieved after 60 days of anaerobic digestion. It is knownthat the decolorization of azo dyes under anaerobic conditionsmay involve both biosorption and reductive cleavage of azobonds. Biosorption is a complex function of the biomasstype, structure, surface properties, surface area and concentra-tion, exposure time to the dyes as well as dye concentration,molecular structure, polarity, and other physicochemical prop-erties. Some researchers established that in the presence of anappropriate carbon source, anaerobic color removal is mainlydue to azo dye reduction [43,44]. In a study conducted withDirect Black 38 at a concentration range of 100e3200 mg/Land 2000 mg/L glucose-COD added as the co-substrate, color

Pigments 78 (2008) 117e130cleavage. This can be supported by a study which was con-ducted with Disperse Blue 79; for low concentrations of this

nd0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00a

so

rb

an

ce at

max, cm

-1)

AR 183 AO 51 RB 4

I. Arslan-Alaton et al. / Dyes adye (>50 mg/L), the use of an external carbon source was notnecessary, since decolorization efficiencies as high as 98%were achieved after 72 h of treatment. However, when the con-centration was higher than 50 mg/L, the use of a co-substrate(114 mg/L sodium acetate) was required to achieve the sametreatment efficiency [43]. Compared with azo dyes, anthraqui-none dyes are less susceptible to reduction and present very lowcolor abatement capacity [49,50]. In an experiment that wasconducted at an initial concentration of 300 mg/L of the anthra-quinone dye Reactive Blue 4 to observe anaerobic color abate-ment, 78% decolorization was obtained after 28 days in the

0.00

0.10

0.20

b

c

0 10 20 30 40 50 60

0 10 20 30 40 50 60

0 10 20 30 40 50 60

Time (days)

Time (days)

Co

lo

r (ab

0

5

10

15

20

25

CH

4 fo

rm

atio

n (m

L C

H4/m

g V

SS

)

CH4 formation Rates (mL CH4(mg VSS d)-1)

CH4 formation Rates (mL CH4(mg VSS d)-1)

AR 183 0.27AO 51 0.46RB 4 0.00Control

AR 183AO 51RB 4Control

0.38

0

5

10

15

20

25

Time (min)

CH

4 fo

rm

atio

n (m

L C

H4/m

g V

SS

)

0.530.500.300.38

AR 183 AO 51 RB 4 Control

AR 183 AO 51 RB 4 Control

Fig. 4. (a) Color abatement and (b) methane production during anaerobic ex-

periments with untreated and (c) Fenton-treated AR 183, AO 51 and RB 5.presence of 8400 mg/LVSS and an external carbon source mix-ture of 750 mg/L dextrine 375 mg/L peptone. In the samestudy, decolorization of 300 mg/L Reactive Blue 19 (anotheranthraquinone dye) was 83% after 28 days [51]. These experi-mental findings were attributed to reductive transformation ofthe dyes [37]. Contrarily, in another study, 68 and 57% color re-movals were obtained for 20 mg/L Reactive Black 5 (a disazodye) and Reactive Blue 19, respectively, after an incubation pe-riod of 16 days and in the presence of 1000 mg/L COD servingas the external carbon source. Decolorization was referred tosorption on bacterial flock material [48]. Generally speaking,when no readily biodegradable co-substrate is provided andno acclimation to the complex substrate is carried out, anaero-bic decolorization can principally be ruled out. Even when theabove given conditions are satisfied it is difficult to make a cor-rect judgment about the dye removal mechanism since abioticcontrol experiments can be very misleading, the structure of thebiosludge changes/deteriorates with biotreatment time leadingto modified biomass surface properties [48]. Hence, in the pres-ent study it was assumed that color removal was mainly due tophysical adsorption (biosorption onto bacterial cell walls).

3.3.2. CH4 formation ratesMethane production was monitored to assess the effect of

untreated and Fenton-treated dyes on anaerobic biomass (meth-anogenic activity). Anaerobic digestion is a multi-step process;methane production is the final, rate-limiting stage of anaerobicdigestion and methane formation rate is a common tool to as-sess toxicity of industrial pollutants [2,49,52]. Fig. 4 presentsmethane production rates in the anaerobic bioreactors fedwith untreated (b) and Fenton-treated (c) textile dyes. FromFig. 4(a)e(c) it is obvious that methane production rate is sig-nificantly faster in the presence of treated textile dyes than forthe control sample, whereas the methane production rate iscomparably slow for sludge digestion in the presence of un-treated textile dyes. No methane production occurred in thesample bearing RB 4. This is in accordance with an investiga-tion of Fontenot et al. [51] who studied the behavior of 500 mg/L aqueous Reactive Blue 4 solution under methanogenicconditions. For an incubation time of 15 days, only 6% CH4production (relative to the control) was observed in that exper-imental work for the sample with Reactive Blue 4. The inhib-itory effect of untreated dyes on methane production can beexplained by the occupation of the methanogenic bacterias ac-tive sites by biosorbed dye molecules [49]. On the other hand,the Fenton-treated dyes had no inhibitory effect on methano-genesis except Fenton-treated RB 4. From Fig. 4(b) and (c) itis also interesting to observe that methane production ratesstarted to level off at different days and that the control samplereached an asymptotic value appreciably later than the treateddyes. The presence of trivalent chromium in AR 183 mighthave had a toxic effect on methanogenic activity. Some relatedpapers point out that heavy metal release from metal-complexdyes during biological and chemical treatment could be themajor reason of the increase in acute toxicity. Osugi et al.

125Pigments 78 (2008) 117e130[53] investigated the toxic effect of the copperephthalocyaninedye Reactive Blue 15 on the photobacteria Vibrio fischeri.

via adsorption and not microbial degradation [56]. However,

well as corresponding controls. As is evident in Fig. 5(b) and(c), nitrate concentration fell from 112 mg/L at the beginningof the anoxic experiment to 0e7.2 mg/L within the first30 min in all bioreactors. Nitrate was completely consumed af-ter 2 h in all bioreactors and initial denitrification rates were

0.00

0.05

0.10

0.15

0.20

0.25

0.30

b

c

0 10 20 30 40 50Time (hours)

Co

lo

r (A

bso

rb

an

ce at

AR 183 AO 51 RB 4

0

20

40

60

80

100

120

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0Time (hours)

NO

3- (m

g/L

)

Denitrification Rates (mg(Lh)-1)AR 183 212AO 51 224RB 4 223Control 212

0

20

40

60

80

100

120

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0Time (hours)

NO

3- (m

g/L

)

AR 183 AO 51 RB 4 Control

AR 183 AO 51 RB 4 Control

Denitrification Rates (mg(Lh)-1)AR 183 210AO 51 224RB 4 217Control 212

Fig. 5. (a) Color abatement and (b) denitrification during anoxic experiments

with untreated and (c) Fenton-treated AR 183, AO 51 and RB 5.

ndit should be kept in mind that in the present study the decolor-ization mechanism for the azo dyes AR 183 and AO 51 couldalso be biosorption. Hence the above explanation is only validif decolorization is a consequence of azo bond reduction forazo dyes and biosorption for anthraquinone dyes. Consideringthat preliminary adsorption accounted for most of the colorremoval and the biomass in all anoxic reactors were coloredafter the experiments were completed, there is definitelymore evidence that under our experimental conditions physicaladsorption onto biomass (biosorption) played a major role incolor removal of the selected textile dyes.

3.4.2. Denitrification ratesPhotoelectrocatalytic treatment caused a rapid increase in theacute toxicity of the dye that was explained by the release ofelemental copper from the dyes structure at a rate of 55%(0.85 mg/L) during its oxidation. In the present study, chro-mium levels in Fenton-treated AR 183 (supernatant) weredetermined as 1.2 mg/L and might probably be the reason forits inhibitory effect on methanogenic bacteria.

The methane production rate in samples containing un-treated, treated dyes as well as control cultures could all befitted to zero-order kinetics. The rate constants are providedin the inset of Fig. 4(b) and (c) and used to evaluate the rela-tive inhibitory effect of the untreated and treated dyes in theforthcoming section.

3.4. Anoxic experiments

3.4.1. ColorColor abatement rates observed for the untreated textile

dyes during anoxic experiments are presented in Fig. 5(a). Pre-liminary adsorption accounted for 62, 92 and 73% decoloriza-tion and total color removal was obtained as 72, 94 and 85%,revealing that the major fraction of absorbance was already re-moved before the denitrification process even started. Providedthat the decolorization mechanism for AR 183 and AO 51 isreduction of the azo bonds, the low color abatement rates ob-served for the two azo dyes compared with RB 4 can be ex-plained as the competition between NO3

and N and the azodye for electron donors. This may be supported with the pre-vious studies where color abatement under anoxic conditionswas investigated. It was observed that NO3

eN acts as an elec-tron acceptor and the presence of an alternative electron accep-tor (namely an azo dye) may retard reductive decolorization[54e56]. Considering that the presence of nitrate could havehindered reductive azo bond cleavage of the two azo dyesbut not the decolorization of the anthraquinone dye in the pres-ent study, it is not surprising that decolorization during denitri-fication was faster for RB 4. In another study carried out with20 mg/L aqueous reactive dye solutions in the presence of 5and 10 mM NO3

eN in a sequential anaerobic/aerobic reactor,the addition of nitrate had no adverse effect on color removalfor the anthraquinone dyes because these dyes were removed

126 I. Arslan-Alaton et al. / Dyes aFig. 5 displays denitrification rates observed during anoxicexperiments in untreated (b) and Fenton-treated (c) samples as0.35

0.40

0.45a

m

ax, cm

-1)

Pigments 78 (2008) 117e130similar to each other (see inset in Fig. 5(b) and (c) for the ini-tial reaction rate constants); from the denitrification rate

treated textile dyes and the corresponding control sample

was a bit lower than that in the samples containing untreateddyes (3700 mg/L in the samples with treated dyes instead of4400 mg/L in the sample with untreated dyes) that directlyaffected the F/M ratio and hence COD abatement rates. How-ever, it should be emphasized that the COD fell down toa range of 17 mg/L (sample with treated RB 4) and 25 mg/L(control and sample with treated AR 183) implying thatglucose degradation was more complete in the presence ofFenton-treated dyes. In all cases, glucose-COD abatement

2

0.000.100.20

b

c

0 1 2 3 4 5 6Time (hours)

Co

lo

r (A

Control AR 183 AO 51 RB 4

Control AR 183 AO 51 RB 4

0

100

200

300

400

500

600

700

0 1 2 3 4 5 6Time (hours)

CO

D (m

g/L

)

Glucose Degradation Rates (h-1)AR 183 3.8AO 51 3.5RB 4 3.7Control 4.9

0

100

200

300

400

500

600

0 1 2 3 4 5 6Time (hours)

CO

D (m

g/L

)

Glucose Degradation Rates (h-1)AR 183 1.30AO 51 0.83RB 4 0.56Control 0.51

Fig. 6. (a) Color abatement and (b) glucose-COD degradation during aerobic

experiments with untreated and (c) Fenton-treated AR 183, AO 51 and RB 5.

nd(VSS 3700 mg/L; F/M 0.13 mg CODo/mg VSS). Com-pared to the COD abatement rates shown in Fig. 6(b), thoseconstants it is clear that the untreated as well as Fenton-treateddyes did not adversely affect anoxic processes.

3.5. Aerobic experiments

3.5.1. ColorFig. 6(a) depicts changes in color during aerobic treatment

of the selected textile dyes. Again, preliminary adsorption ontobiomass was also examined (not shown data) and found as 6,93 and 41%, whereas overall color removal efficiencies wereobtained as 22, 96 and 84% for AR 183, AO 51 and RB 4, re-spectively. It is known that azo-, nitro-, or sulfo-substitutionsare not readily biodegradable under aerobic conditions [57].Hence in the present study, color removal is expected to bea consequence of biosorption onto activated sludge. Buitronet al. [58] investigated the aerobic degradation of 25 and50 mg/L Acid Red 151 solution in a sequencing batch biofilterpacked with porous volcanic rock. In this stuy, Acid Red 151was used as the sole carbon and energy source for microorgan-isms. After 24 h, 50% removal of the initial absorbance wasobserved; however, the removal mechanism was adsorption(onto the packing material) and biosorption (onto attached mi-croorganisms). Pourbabaee et al. [59] studied the decoloriza-tion of 300 mg/L Terasil Black (a fiber disperse dye) underaerobic conditions in the presence of an exogenous carbonsource (e.g. glucose and yeast extract). Decolorization of theeffluent containing Terasil Black dye depended upon the pres-ence of external carbon and energy sources. Nearly 60% of theoriginal color disappeared after 30 h of incubation at a temper-ature of 30 C in the presence of 10 g/L glucose and 5 g/Lyeast extract, whereas only 10% decolorization occurred intheir absence. Disperse dyes usually agglomerate fast andtend to precipitate out of the solution bulk easily. In our study,activated sludge biomass preferred to consume the easilydegradable carbon source (glucose) instead of the carbon beingincorporated in the complex dye molecules.

3.5.2. Glucose degradation ratesFig. 6(b) shows glucose-COD degradation rates for un-

treated textile dyes and the control sample (operating condi-tions: VSS 4400 mg/L; F/M 0.12 mg CODo/mg VSS).From Fig. 6(b) is can be seen that glucose degradation waspractically complete within 30 min due to glucose exhaustionfor all studied samples. The final CODs that were reached after6 h aerobic biotreatment ranged between 37 mg/L (control)and 65 mg/L (sample containing glucose plus AR 183); the re-sidual CODs were thought to be metabolic end products ofglucose plus textile dyes that were not biosorbed onto acti-vated sludge. This result also explains why the remainingCOD was slightly higher for the sample bearing AR 183,which was at least sorbed onto aerobic cell biomass.Fig. 6(c) gives glucose-COD degradation rates for Fenton-

I. Arslan-Alaton et al. / Dyes agiven in Fig. 6(c) were significantly slower because the VSSconcentration in the samples bearing Fenton-treated dyes0.300.400.500.600.700.800.901.00

a

bso

rb

an

ce at

max, cm

-1)

AR 183 AO 51 RB 4

127Pigments 78 (2008) 117e130rates followed pseudo-first order kinetics (R of the kineticfit> 0.95) with respect to COD and the corresponding reaction

rates183,lated

on the basis of the reaction rate constants derived from the an-

the textile dye sample) 100/(methane formation or denitrification or glucose

ndaerobic (see Fig. 4), anoxic (Fig. 5) and aerobic (Fig. 6) exper-iments relative to the corresponding control reactors. FromTable 4 it can be inferred that treated dyes exhibited no or ap-preciably less inhibitory effect on biomass than the untreateddyes implying that Fenton oxidation is an ecotoxicologicallysafe treatment option and can potentially be applied as a pre-periments) and glucose degradation (aerobic experiments)obtained for untreated and Fenton-treated textile dyes ARAO 51 and RB 4. The percent inhibition values were calcurate coefficients are listed in the inset of Fig. 6(b) and (c).From the rate constants it can be concluded that untreatedtextile dyes had a slightly inhibitory and the Fenton-treateddyes a significantly accelerating (providing more degradablesubstrate to the bioreactor) effect on glucose degradationwith activated sludge.

3.6. Evaluation of the inhibitory effect on anaerobic,anoxic and aerobic treatment processes

Table 4 summarizes percent inhibition (I, %) of the methaneproduction (anaerobic experiments), denitrification (anoxic ex-

degradation rate constant in the control sample).Table 4

Percent inhibition of CH4 formation ICH4 , denitrification INO3 and glucose-COD (ICOD) abatement rates calculated for untreated and Fenton pre-treated

(reaction conditions: Fe2 4 mM; H2O2 20 mM; pHo 3; t 10 min)AR 183, AO 51 and RB 4 (100 mg/L aqueous textile dye solutions) relative

to the corresponding control experiments

Textile

dye

Anaerobic ICH4 (%)a Anoxic INO

3(%)a Aerobic ICOD (%)

a

Untreated Treated Untreated Treated Untreated Treated

AR 183 29 0 0 0 23 0

AO 51 0 0 0 0 29 0

RB 4 100 21 0 0 25 0

a Percent inhibition was calculated as follows: I (%) (methane formationor denitrification or glucose degradation rate constant in the control samplemethane formation or denitrification or glucose degradation rate constant in

128 I. Arslan-Alaton et al. / Dyes atreatment stage for fast/complete color and partial organic car-bon removal prior to anaerobic, anoxic and aerobic processes.

4. Conclusions and recommendations

Textile dyes are known to resist conventional physicochem-ical (adsorption, coagulation) and aerobic biological treatmentdue to their high degree of polarity and complex molecularstructure. As a consequence, textile dyes tend to sorb onactivated sludge in aerobic biotreatment systems and on soilsediments in receiving water bodies, causing significant envi-ronmental, ecotoxicological and ecological problems. Withthese important issues in mind, the present study aimed at treat-ing three commercially important textile dyes (Acid Red 183,AR 183; Acid Orange 51, AO 51; and Reactive Blue 4, RB 4)with Fentons reagent to evaluate the effect of untreated andFenton-treated textile dyes on aerobic, anoxic and anaerobicprocesses. Our results have demonstrated that during Fentontreatment of the selected dyes decolorization was completewithin minutes and accompanied with appreciable COD andDOC removals. Acetate was detected as the major stable oxida-tion product and its concentration depended upon the dye type.Nitrate was expected due to the amino, nitro, azo, and triazynylcontent of the investigated textile dyes, but did not form duringFenton treatment at significant levels (its concentration re-mained at the detection limit of the analytical equipment).Our experimental findings have also indicated that textiledyes have less inhibitory effect on anaerobic (methanogenesis)and aerobic (glucose abatement) processes than the untreated(original) ones when subjected to Fenton treatment under opti-mized reaction conditions. The untreated and Fenton-treatedtextile dyes did not exert any inhibitory effect on denitrification.On the other hand, our study indicated that decolorization viabiological processes was mainly due to sorption onto biomassand only efficient for the disazo dye AO 51. Nevertheless, Fen-ton process can be recommended for complete, oxidative colorand partial organic carbon removal; a feasible solution to meetcolor discharge standards, reduce the organic load associatedwith the dyes in dyehouse effluent without causing any inhibi-tory effect on anaerobic, anoxic and aerobic processes.

Acknowledgements

The present work was financially supported by TUBA (TheTurkish Academy of Sciences) under the Young ScientistsScholarship Program and conducted in the laboratories ofTechnical University of Denmark (Fenton, anaerobic, anoxicexperiments) and Istanbul Technical University (Fenton, aero-bic experiments).

References

[1] Manu B, Chaudhari S. Anaerobic decolorisation of simulated textile

wastewater containing azo dyes. Bioresource Technology 2002;82:

225e31.

[2] Bell J, Plumb JJ, Buckley CA, Stuckey DC. Treatment and decolorization

of dyes in an anaerobic baffled reactor. Journal of Environmental Engi-

neering 2000;126:1026e32.

[3] Lee YH, Pavlostathis SG. Decolorization and toxicity of anthraquinone

textile dyes under methanogenic conditions. Water Research 2004;38:

1838e52.

[4] Novotny C, Dias N, Kapanen A, Malachova K, Vandrovcova M,

Itarvaara M, et al. Comparative use of bacterial, algal and protozoan tests

to study toxicity of azo and anthraquinone dyes. Chemosphere 2006;63:

1436e42.

[5] Slokar YM, Marechal M. Methods of discoloration of textile wastewa-

ters. Dyes and Pigments 1997;37:335e56.[6] Hao OJ, Kim H, Chiang PC. Decolorization of wastewater. Critical Re-

views in Environmental Science and Technology 2000;30:449e505.

[7] McCallum JEB, Madison SA, Alkan S, Depinto RL, Wahl RUR.

Analytical studies on the oxidative degradation of the reactive textile

dye Uniblue A. Environmental Science and Technology 2000;34:

5157e64.

[8] Aplin R, Waite T. Comparison of three advanced oxidation processes for

degradation of textile dyes. Water Science and Technology 2000;42:

345e54.

[9] Arslan-Alaton I, Akmehmet-Balcioglu I, Bahnemann DW. Advanced ox-

Pigments 78 (2008) 117e130idation of a reactive dyebath effluent: comparison of O3, H2O2/UV-C and

TiO2/UV-A processes. Water Research 2002;36:1143e54.

nd[10] Joseph JM, Destaillats H, Hung H, Hoffmann MR. The sonochemical

degradation of azobenzene and related azo dyes: rate enhancements via

Fentons reactions. Journal of Physical Chemistry 2000;A104:301e7.

[11] Chamarro E, Marco A, Esplugas S. Use of Fenton reagent to improve

organic chemical biodegradability. Water Research 2001;35:1047e51.[12] Fongsatitkul P, Elefsiniotis P, Yamasmit A, Yamasmit N. Use of sequenc-

ing-batch reactors and Fentons reagent to treat a wastewater from a tex-

tile industry. Biochemical Engineering Journal 2004;21:213e20.[13] Kim TH, Park C, Yang J, Kim S. Comparison of disperse and reactive

dye removals by chemical coagulation and Fenton oxidation. Journal

of Hazardous Materials 2004;B112:95e103.

[14] Meric S, Kaptan D, Olmez T. Color and COD removal from wastewater

containing Reactive Black 5 using Fentons oxidation process. Chemo-

sphere 2004;54:435e41.

[15] Hsueh CL, Huang YH, Wang CC, Chen CY. Degradation of azo dyes us-

ing low iron concentration of Fenton and Fenton-like system. Chemo-

sphere 2005;58:1409e14.

[16] Arslan-Alaton I, Teksoy S. Acid dyebath effluent pretreatment using Fen-

tons reagent: process optimization, reaction kinetics and effects on acute

toxicity. Dyes and Pigments 2007;73:31e9.

[17] Lucas MS, Dias AA, Sampaio A, Amaral C, Peres J. Degradation of a tex-

tile reactive azo dye by a combined chemicalebiological process:

Fentons reagent-yeast. Water Research 2007;41:1103e9.[18] Velegraki T, Poulios I, Charalabaki M, Kalogerakis N, Samaras P,

Mantzavinos D. Photocatalytic and sonolytic oxidation of acid orange

7 in aqueous solution. Applied Catalysis B 2006;62:159e68.

[19] Garcia-Montano J, Torrades F, Garcia-Hortal JA, Domenech X, Peral J.

Combining photo-Fenton process with aerobic sequencing batch reactor

for commercial hetero-bireactive dye removal. Applied Catalysis B

2006;67:86e92.[20] Garcia-Montano J, Torrades F, Garcia-Hortal JA, Domenech X, Peral J.

Degradation of Procion Red H7B reactive dye by coupling a photo-

Fenton system with a sequencing batch reactor. Journal of Hazardous

Materials 2006;B134:220e9.[21] Tantak N, Chaudari S. Degradation of azo dyes by sequential Fentons

oxidation and aerobic biological treatment. Journal of Hazardous Mate-

rials 2006;B136:698e705.

[22] Lodha B, Chaudari S. Optimization of Fenton-biological treatment

scheme for the treatment of aqueous dye solutions. Journal of Hazardous

Materials 2007;B148:459e66.

[23] ONeill C, Hawkes FR, Hawkes D, Lourenco ND, Pinheiro HM,

Delee W. Review colour in textile effluents e sources, measurement, dis-charge consents and simulation: a review. Journal of Chemical Technol-

ogy and Biotechnology 1999;74:1009e18.

[24] Akin Tekstil. Private communication; 2002.

[25] Pisa Tekstil. Private communication; 2004.

[26] Turbo Tekstil. Private communication; 2006.

[27] Danish Standard. Water examinationedetermination of chemical oxygen

demand in water with dichromate. DS 217; 1991 [in Danish].

[28] ISO 6060. Water quality-determination of the chemical oxygen demand.

2nd ed., vol. 147. Geneva: ISO 6060/TC; 1989.

[29] APHA-AWWA-WPCF. Standard methods for the examination of water

and wastewater. 20th ed. Washington, DC: American Public Health

Association; 1998.

[30] Vajnhandl S, Marechal AM. Case study of the sonochemical decoloriza-

tion of textile azo dye Reactive Black 5. Journal of Hazardous Materials

2006;136:698e705.

[31] Duran A, Monteagudo JM, Mohedano M. Neural networks simulation of

photo-Fenton degradation of Reactive Blue 4. Applied Catalysis B Envi-

ronmental 2006;65:127e34.[32] Ramirez JH, Costa CA, Maderia LM. Experimental design to optimize

the degradation of synthetic dye orange II using Fentons reagent. Catal-

ysis Today 2005;107:68e76.

[33] Malik PK, Saha SK. Oxidation of direct dyes with hydrogen peroxide

using ferrous ion as catalyst. Separation and Purification Technology

2003;31:241e50.

I. Arslan-Alaton et al. / Dyes a[34] Neamtu M, Yediler A, Siminiceanu I, Kettrup A. Oxidation of commer-

cial reactive azo dye aqueous solutions by the Photo-Fenton and Fenton-like processes. Journal of Photochemistry and Photobiology A: Chemis-

try 2003;161:893e9.

[35] Neamtu M, Yediler A, Siminiceanu I, Macoreanu M, Kettrup A. Decol-

orization of Disperse Red 354 azo dye in water by several oxidation pro-

cesses e a comparative study. Dyes and Pigments 2004;60:61e8.[36] Song S, Ying H, He Z, Chen J. Mechanism of decolorization and degra-

dation of CI Direct Red 23 by ozonation combined with sonolysis. Che-

mosphere 2007;66:1782e8.[37] Jozwiak WK, Mitros M, Kaluzina-Czaplinska J, Tosik R. Oxidative de-

composition of Acid Brown 159 dye in aqueous solution by H2O2/

Fe2 and ozone with GC/MS analysis. Dyes and Pigments 2007;74:9e16.

[38] Zhang F, Yediler A, Liang X. Decomposition pathways and reaction

intermediate formation of the purified, hydrolyzed azo reactive dye

C.I. Reactive Red 120 during ozonation. Chemosphere 2007;67:

712e7.[39] Matsui M, Kobayashi K, Shibata K, Takase Y. Ozonation of dyes (II) e

ozone treatment of 4-phenyl-azo-1-naphthol. Journal of the Society of

Dyers and Colourists 1981;97:210e3.[40] Gahr F, Hermanutz F, Oppermann W. Ozonation: an important technique

to comply with new German laws for textile wastewater treatment. Water

Science and Technology 1994;30:255e63.

[41] Chiron S, Rodriguez A, Fernandes-Alba A. Application of gas and liquid

chromatographyemass spectrometry to the evaluation of pirimiphos

methyl degradation products in industrial water and ozone treatment.

Journal of Chromatography A 1998;823:97e107.

[42] Gutowska A, Kaluzna-Czaplinska J, Jozwiak WK. Degradation mecha-

nism of Reactive Orange 113 dye by H2O2/Fe2 and ozone in aqueous

solution. Dyes and Pigments 2007;74:41e6.

[43] Cruz A, Buitron G. Biodegradation of Disperse Blue 79 using sequenced

anaerobic/aerobic biofilters. Water Science and Technology 2001;44:

159e66.

[44] Pinheiro HM, Touraud E, Thomas O. Aromatic amines from azo dye re-

duction: status review with emphasis on direct UV spectrophotometric

detection in textile industry wastewaters. Dyes and Pigments 2004;61:

121e39.

[45] Sponza DT, Isik M. Toxicity and intermediates of C.I. Direct Red 28 dye

through sequential anaerobic/aerobic treatment. Process Biochemistry

2005;40:2735e44.

[46] Albuquerque MGE, Lopes AT, Serralheiro ML, Novais JM,

Pinheiro HM. Biological sulphate reduction and redox mediator effects

on azo dye decolorization in anaerobiceaerobic sequencing batch reac-tors. Enzyme and Microbial Technology 2005;36:790e9.

[47] Maas R, Chaudhari S. Adsorption and biological decolorization of azo

dye Reactive Red 2 in semicontinuous anaerobic reactors. Process Bio-

chemistry 2005;40:699e705.

[48] Luangdilok W, Passwad T. Effect of chemical structures of reactive dyes

on color removal by anaerobiceaerobic process. Water Science and

Technology 2000;42:377e82.[49] Dos Santos AB, Bisschops IAE, Cervantes FJ, van Lier JB. The transfor-

mation and toxicity of anthraquinone dyes during thermophilic (55 C)and mesophilic (30 C) anaerobic treatments. Journal of Biotechnology2005;115:345e53.

[50] Epolito WJ, Lee YH, Bottomley LA, Pavlostathis SG. Characterization

of the textile anthraquinone dye Reactive Blue 4. Dyes and Pigments

2005;67:35e46.[51] Fontenot EJ, Beydilli MI, Lee YH, Pavlostathis SG. Kinetics and inhibi-

tion during the decolorization of reactive anthraquinone dyes under

methanogenic conditions. Water Science and Technology 2002;

45:105e11.[52] Donlon BA, Razo-Flores E, Luijten M, Swarts H, Lettinga G, Field JA.

Detoxification and partial mineralization of the azo dye Mordant Orange

1 in a continuous upflow anaerobic sludge-blanket reactor. Applied Mi-

crobiology and Biotechnology 1997;47:83e90.[53] Osugi ME, Umbuzeiro GA, De Castro FJV, Zanoni VBZ. Photoelectro-

catalytic oxidation of Remazol Turquoise Blue and toxicological assess-

129Pigments 78 (2008) 117e130ment of its oxidation products. Journal of Hazardous Materials 2006;137:

871e7.

[54] Carliell CM, Barclay SJ, Naidoo N, Buckley CA, Muholland DA,

Senior E. Microbial decolorization of a reactive azo dye under anaerobic

conditions. Water South Africa 1995;21:61e9.

[55] Laurenco ND, Novais JM, Pinheiro HM. Reactive textile dye colour re-

moval in a sequencing batch reactor. Water Science and Technology

2000;42:321e8.

[56] Panswad T, Luangdilok W. Decolorization of reactive dyes with different

molecular structures under different environmental conditions. Water Re-

search 2000;34:4177e84.

[57] Zimmerman FJ, Diddams DG. Zimmermann process and its applications

in pulp and paper industry. Tappi 1960;43:710e5.

[58] Buitron G, Quezada M, Moreno G. Aerobic degradation of the azo dye

Acid Red 151 in a sequencing batch biofilter. Bioresource Technology

2004;92:143e9.[59] Pourbabaee AA, Malekzadeh F, Sarbolouki MN, Najafi F. Aerobic decol-

orization and detoxification of a disperse dye in textile effluent by a new

isolate of Bacillus sp. Biotechnology and Bioengineering 2006;93:631e5.

130 I. Arslan-Alaton et al. / Dyes and Pigments 78 (2008) 117e130

Advanced oxidation of acid and reactive dyes: Effect of Fenton treatment on aerobic, anoxic and anaerobic processesIntroductionMaterials and methodsTextile dyesFenton treatmentAnaerobic experimentsAnoxic experimentsAerobic experimentsAnalytical tools

Results and discussionFenton experiments: establishment of optimum working conditionsFenton treatmentColorCODDOCIdentification of final (stable) oxidation products

Anaerobic experimentsColorCH4 formation rates

Anoxic experimentsColorDenitrification rates

Aerobic experimentsColorGlucose degradation rates

Evaluation of the inhibitory effect on anaerobic, anoxic and aerobic treatment processes

Conclusions and recommendationsAcknowledgementsReferences